BACKGROUND: Glenoid component loosening after total shoulder arthroplasty (TSA) may occur if retroversion is not corrected to <10°. However, accurately measuring postoperative glenoid component version has been difficult without postoperative computed tomography (CT), adding cost and radiation exposure outside of the standard radiographic follow-up. We present a new method to assess glenoid component version after TSA using only routine preoperative CT and postoperative radiographs (x-rays). METHODS: Preoperative glenoid version was measured using established methods with an axillary x-ray, 2-dimensional CT, and Glenosys software (Imascap, Plouzané, France). Postoperative glenoid component version and inclination were measured for 61 TSA patients using Mimics software (Materialise, Leuven, Belgium) with preoperative CT and postoperative x-rays. Four patients also had postoperative CTs. Glenoid implantation and imaging were performed on 14 cadavers, allowing validation of results against the gold standard postoperative CT glenoid retroversion measurement. RESULTS: Compared with the gold standard, retroversion and inclination measurement error was 2° ± 1° and 2° ± 1°, respectively. Average postoperative version correction was 6° ± 7°, with 35 of 61 patients (57%) corrected to <10° of retroversion. Correlation between preoperative version measurement methods was good to very good, except on the axillary x-ray. Patients not corrected to <10° of retroversion had significantly higher preoperative retroversion (14° ± 6°) than those corrected to <10° (6° ± 7°; P < .00001). CONCLUSIONS: Glenoid component retroversion after TSA can be accurately measured with a method using only routine preoperative CT and postoperative x-rays, validated to within 1.9° of the gold standard postoperative CT measurement. Future studies using this method may correlate glenoid retroversion correction with glenoid component longevity to help optimize shoulder arthroplasty outcomes.
BACKGROUND: Glenoid component loosening after total shoulder arthroplasty (TSA) may occur if retroversion is not corrected to <10°. However, accurately measuring postoperative glenoid component version has been difficult without postoperative computed tomography (CT), adding cost and radiation exposure outside of the standard radiographic follow-up. We present a new method to assess glenoid component version after TSA using only routine preoperative CT and postoperative radiographs (x-rays). METHODS: Preoperative glenoid version was measured using established methods with an axillary x-ray, 2-dimensional CT, and Glenosys software (Imascap, Plouzané, France). Postoperative glenoid component version and inclination were measured for 61 TSA patients using Mimics software (Materialise, Leuven, Belgium) with preoperative CT and postoperative x-rays. Four patients also had postoperative CTs. Glenoid implantation and imaging were performed on 14 cadavers, allowing validation of results against the gold standard postoperative CT glenoid retroversion measurement. RESULTS: Compared with the gold standard, retroversion and inclination measurement error was 2° ± 1° and 2° ± 1°, respectively. Average postoperative version correction was 6° ± 7°, with 35 of 61 patients (57%) corrected to <10° of retroversion. Correlation between preoperative version measurement methods was good to very good, except on the axillary x-ray. Patients not corrected to <10° of retroversion had significantly higher preoperative retroversion (14° ± 6°) than those corrected to <10° (6° ± 7°; P < .00001). CONCLUSIONS: Glenoid component retroversion after TSA can be accurately measured with a method using only routine preoperative CT and postoperative x-rays, validated to within 1.9° of the gold standard postoperative CT measurement. Future studies using this method may correlate glenoid retroversion correction with glenoid component longevity to help optimize shoulder arthroplasty outcomes.